Light and strange quark masses with N f = 2 + 1 N_{f}=2+1 Wilson fermions
Alejandro Sáez · 2026
Read Light and strange quark masses with N f = 2 + 1 N_{f}=2+1 Wilson fermions by Alejandro Sáez (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-26623.txt.
Introduction
Herdoíza and co-authors report progress on computing light and strange quark masses in lattice QCD. The run uses three dynamical flavours with improved Wilson fermions on CLS ensembles. Bare masses come from the simulations, then are renormalised without perturbation theory in the Schrödinger functional scheme. The scheme spans a wide range of scales before meeting perturbative results. Five lattice spacings reach down to about 0.038 femtometres, with pion masses at the physical value. The spread lets them test cutoff and chiral effects before extrapolating.
Catalog record
| CREATOR | Alejandro Sáez |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 17845 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.26623
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Light and strange quark masses with N f = 2 + 1 N_{f}=2+1 Wilson fermions Alejandro Sáez Abstract We report on the status of an update of our collaboration’s previous computation [ 4 ] of light and strange quark masses in QCD with N f = 2 + 1 N_{f}=2+1 dynamical flavours. Bare quark masses are extracted from CLS ensembles, using O ( a ) O(a) -improved Wilson fermions, and the mass renormalization is performed non-perturbatively in the Schrödinger functional scheme over a wide range of scales to make safe contact with perturbation theory. Results for five lattice spacings, down to a ∼ 0.038 fm a\sim 0.038\textrm{ fm} , and pion masses reaching the physical value are included in the analysis. This allows for the exploration of different models for cutoff and chiral effects, and a controlled extrapolation to the physical point. IFT-UAM/CSIC-26-41 [a]Fernando P. Panadero 1 Introduction Quark masses are fundamental parameters of the Standard Model. A first-principles determination of these quantities requires a consistent non-perturbative formulation of QCD. We present a preliminary update on [ 4 ] for the determination of the light and strange quark masses, in a framework where the up and down quarks are mass-degenerate and QED effects are ignored, known as N_{f}=2+1 QCD. The status for the determination of quark masses in Lattice QCD with this setup is summarized in the Flavour Lattice Averaging Group (FLAG) Review [ 1 ] . The current values for the average are \overline{m}_{ud}^{\rm\scriptscriptstyle\overline{MS}}(2\textrm{GeV})=3.387(39)\textrm{MeV},\qquad\overline{m}_{s}^{\overline{\rm\scriptscriptstyle MS}}(2\textrm{GeV})=92.4(1.0)\textrm{MeV}. (1) To determine the physical values of quark masses, we employ large-scale simulations with N_{f}=2+1 dynamical flavours produced by the Coordinated Lattice Simulations (CLS) effort [ 6 ] . These ensembles employ a tree-level Symanzik-improved gauge action and non-perturbatively improved Wilson fermions in the sea sector. We consider a set of ensembles with five lattice spacings in the range 0.085\text{-}0.038\textrm{ fm} and pion masses from 420\textrm{ MeV} down to the physical pion mass. The quark masses extracted from these simulations are bare quantities and have to be renormalized. For this, we use the non-perturbative renormalization and Renormalization Group (RG) running results from the ALPHA Collaboration [ 9 ] , based on a Schrödinger functional (SF) finite volume renormalization scheme. This allows to run from a hadronic scale, where the quark masses obtained in the large volume simulations are extracted and renormalized, to a very high energy perturbative scale, where one can make safe contact with perturbation theory. The main update with respect to [ 4 ] is the extension of the dataset with six new ensembles that include one additional lattice spacing and physical kinematics ensembles, allowing for a better controlled chiral-continuum extrapolation to the physical point, as well as a new scale setting determination [ 8 ] . 2 Lattice setup and ensembles In this section, we summarize the theoretical framework used to define renormalized quark masses and give some details of the ensembles entering the analysis. 2.1 Theoretical framework The main observables for the computation of quark masses are the pseudoscalar-pseudoscalar and axial-pseudoscalar bare correlation functions f_{P}^{ij}(x_{0},y_{0})=-\frac{a^{6}}{L^{3}}\sum_{\vec{x},\vec{y}}\langle P^{ij}(x_{0},\vec{x})P^{ji}(y_{0},\vec{y})\rangle,\qquad f_{A}^{ij}(x_{0},y_{0})=-\frac{a^{6}}{L^{3}}\sum_{\vec{x},\vec{y}}\langle A_{0}^{ij}(x_{0},\vec{x})P^{ji}(y_{0},\vec{y})\rangle, (2) where A_{\mu}^{ij}(x)={\overline{\psi}}^{i}(x)\gamma_{\mu}\gamma_{5}\psi^{j}(x) , P^{ij}(x)={\overline{\psi}}^{i}(x)\gamma_{5}\psi^{j}(x) and i,j=1,2,3 label quark flavour. By imposing the PCAC Ward identity \partial_{\mu}A_{\mu}^{ij}(x)=2m_{ij}P^{ij}(x)=(m_{i}+m_{j})P^{ij}(x),\qquad\text{with}\qquad i\neq j, (3) we can define the bare PCAC quark mass by using the pseudoscalar density to probe this relation. m_{ij}(x_{0})=\frac{\frac{1}{2}(\partial_{0}+\partial^{*}_{0})f_{A}^{ij}(x_{0},y_{0})+c_{A}a\partial_{0}\partial^{*}_{0}f^{ij}_{P}(x_{0},y_{0})}{2f_{P}^{ij}(x_{0},y_{0})}, (4) where the mass independent improvement coefficient c_{A} is non-perturbatively determined in [ 7 ] . The renormalized quark masses are then defined as a\overline{m}_{ij}(g_{0},a\mu)=\frac{Z_{\rm\scriptscriptstyle A}(g_{0}^{2})}{Z_{\rm\scriptscriptstyle P}(g_{0}^{2},a\mu)}am_{ij}\left[1+(\tilde{b}_{A}-\tilde{b}_{P})am_{ij}\right], (5) where the renormalization factors Z_{\rm\scriptscriptstyle P}(g_{0},a\mu) [ 9 ] , Z_{\rm\scriptscriptstyle A}(g_{0}^{2}) [ 10 ] and the improvement coefficient (\tilde{b}_{\rm\scriptscriptstyle A}-\tilde{b}_{\rm\scriptscriptstyle P}) [ 11 ] have been computed in previous works. In this expression, we are neglecting improvement terms proportional to \textrm{Tr}(aM_{q}) . These terms are expected to have a small effect with our current precision as they are \textrm{O}(g_{0}^{4}) in perturbation theory, see [ 8 ] for more details. After the chiral-continuum extrapolation, this expresion yields the renormalized value of the quark masses in the Schrödinger Functional scheme at \mu_{\rm\scriptscriptstyle had}=233(8)\rm{MeV} . We can then convert this to the RGI masses with the RG factor [ 9 ] \frac{M_{i}^{\rm\scriptscriptstyle RGI}}{\overline{m}_{i}(\mu_{\rm\scriptscriptstyle had})}=0.9148(88). (6) As this is an RGI and scheme-independent quantity, it can later be used to extract the quark mass in any other renormalization scheme via M^{\rm\scriptscriptstyle RGI}_{i}=\overline{m}_{i}(\mu)\left(2b_{0}\overline{g}^{2}(\mu)\right)^{-\frac{d_{0}}{2b_{0}}}\exp\left\{-\int_{0}^{\overline{g}(\mu)}dg\left(\frac{\tau(g)}{\beta(g)}-\frac{d_{0}}{b_{0}g}\right)\right\}. (7) Where \beta(g),\tau(g) are respectively the beta function and the quark mass anomalous dimension. In practice, we use the flow scale t_{0} [ 16 ] to scale both hadron and quark masses: \phi_{2}=8t_{0}m_{\pi}^{2},\qquad\phi_{4}=8t_{0}(m_{K}^{2}+\frac{1}{2}m_{\pi}^{2}),\qquad\phi_{K}=8t_{0}m_{K}^{2} (8) \phi_{12}=\sqrt{8t_{0}}\text{ }\overline{m}_{12},\qquad\phi_{13}=\sqrt{8t_{0}}\text{ }\overline{m}_{13}. (9) 2.2 Simulation Setup We employ a subset of the N_{f}=2+1 CLS ensembles [ 6 ] , generated with O(a) -improved Wilson fermion and Lüscher-Weisz action. We use the \textrm{Tr}(M_{q})=\textrm{const} trajectory of CLS, ensuring the lattice spacing is kept fixed, up to O(a^{2}) , for a fixed value of the bare coupling. A list of the ensembles is given in Table 1 . Name \beta a [fm] T/a L/a m_{\pi}[\textrm{MeV}] m_{K}[\textrm{MeV}] m_{\pi}L N_{cnfg} H101 3.4 0.085 96 32 426 426 5.8 2010 H102 96 32 360 446 4.9 2005 H105 96 32 286 470 3.9 2826 H400 3.46 0.075 96 32 429 429 5.2 1045 D450 128 64 221 483 5.4 250 N202 3.55 0.063 128 48 418 418 6.5 899 N203 128 48 350 448 5.4 1543 N200 128 48 288 470 4.4 1712 D200 128 64 204 488 4.2 2001 E250 192 96 131 497 4.0 100 N300 3.70 0.049 128 48 427 427 5.1 1521 N302 128 48 350 457 4.2 2201 J303 192 64 261 481 4.1 1073 E300 192 96 177 499 4.2 227 J500 3.85 0.038 192 64 418 418 5.2 422 J501 192 64 339 453 4.3 787 Table 1: Set of N_{f}=2+1 ensembles entering this analysis. All ensembles have open boundary conditions in the time direction, with the exception of E250 and D450 that have (anti-)periodic boundary conditions. Pion and kaon masses and lattice spacings in physical units are approximate values, see [ 8 ] for details. We follow the same strategy as in Refs. [ 8 , 5 ] to fix a constant value of \phi_{4} , computing the mass derivative of primary observables with respect to quark masses and shifting them to a target value of \phi_{4} . The physical values of \phi_{2} and \phi_{4} are determined in [ 8 ] using the so-called Edinburgh Consensus to define isoQCD [ 1 ] . In figure 1 we illustrate the ensembles used in this analysis as a function of the lattice spacing and the pion mass. Note that, because of the chosen chiral trajectory of constant \textrm{Tr}(M_{q}) , the kaon has the same mass as the pion at the SU(3)- symmetric point, where the pion is heavier, and becomes heavier at lighter pion masses. This feature of the chiral trajectory can also be seen in the right plot of figure 1 , where we show the value for the renormalized average quark masses as a function of the pion mass for various lattice spacings. Figure 1: Left: values of the lattice spacing and pion mass for the ensembles used in this work. The new ensembles with respect to [ 4 ] are highlighted. Right: Values of the light-light (